This is reposted from a Q&A we did in January 2023. Check out our Slack for more.
In an exciting interview, we had the privilege of chatting with Austin Che, one of the brilliant minds and founders behind Ginkgo Bioworks. We dive into the promising prospects of synthetic biology, the intricacies of developing lab automation software, Austin's unique journey as a software engineer in the field of biology, and much more!
Question 1:
Nicholas: Welcome @austin and thanks so much for being here with us. i’m super excited for this q&a — ginkgo is a trailblazer in this space and hopefully we can all learn something! i’ll get us started off
as always, please jump in with your questions!
Question 2:
Nicholas: Tell us about your path to ginkgo. how did you get interested in software & synbio?
Austin: I had been programming since when i was around 7 and it was an easy path into cs at stanford where i did my undergrad. but for grad school, i wanted to get into a different area where i could apply my software skills to something other than typical software jobs. i applied to a wide variety of programs and by chance, i met tom knight when i was visiting mit and he introduced me to the concept that we could program biology. that's where all the founders of ginkgo met and the roots of the company began.
Question 3:
Nicholas: You’ve been at ginkgo since the beginning. what roles have you played there?
Austin: I've worked on the software, worked in the lab, managed a number of g&a functions, and generally do what founders do, which is whatever is needed at any given time.
Question 4:
Nicholas: What types of software does ginkgo build internally?
Austin: We have software around tracking samples like lims, we have automation control software to run a variety of workflows, we have software to help us design dna, software for data analysis, software for tracking all the genomes and other experiments we're doing, and probably lots of other things also.
Question 5:
Nicholas: A lot of commercial software is targeted towards the biopharma industry. what pieces of software have you been able to adapt to fit your use cases?
Austin: We use a lot of off the shelf software from project management tools, software development tools, instrument software, sequence tools like geneious, and software for running data analysis pipelines.
Question 6:
Nicholas: How have you built a culture that encourages programmers to interface with scientists effectively?
Austin: Early on, many of our programmers came from the biology side and were familiar with the lab. more recently with hiring more generalist programmers, we've needed to push more education about what happens in the lab to them. for example, many of our programmers take a class through biobuilder where they learn about synbio and do some experiments themselves. hiring people fluent on both software and biology to be the interface has also been helpful. we also have created specific teams of people that come from a science background to focus on the more science related projects vs the software projects that may be more infrastructure focused.
Nicholas: Do you have any thoughts on how we can encourage this overlap more broadly? any resources that you have developed that are effective?
Austin: The issue of people being fluent in both software and biology will go away in time. software is becoming easier for biologists to pick up. but in general, good software engineers in any domain needs to understand the customers and how they will be using the software. so this isn't specific to our industry. there is some self selection for programmers who care about biology.
Question 7:
Joseph: Can you give us more details on how ginkgo bioworks supports the vaccine supply chain?
Austin: We worked with aldevron to optimize vaccinia capping enzyme which is used in manufacturing mrna vaccines. our process was over 10x more efficient.
Question 8:
Cory: I would like to hear more about the the relationship you see between your cs background and programming biology. in what ways do the approaches you learned in cs apply to biological engineering?
Austin: The key is to take an engineering approach to biology rather than a scientific approach. the scientist aims to study what is there and the engineer tries to reduce the complexity of it. cs is good at teaching abstraction and hiding complexity and there's a lot of complexity to hide in biology.
Cory: Thanks for the response, approaching synthetic biology from the perspective of reducing complexity is very insightful. i've personally been trying to find people who can apply engineering concepts to designing biological platforms, and it seems to me overlap between the two fields is constantly growing.
Question 9:
Simon: Thanks for doing the q&a with us! what are the promising areas you see for synthetic biology outside of therapeutics?
Austin: Climate change for sure. a carbon problem on a global scale cries for a biological solution.
Simon: Exciting! in your view, what is the notable project demonstrating traction for this, at ginkgo or elsewhere?
Austin: Living carbon is engineering plants to capture more carbon. plants are about 80% of the biomass of the planet and something like this takes advantage of the intrinsic potential scale of biology.
Question 10:
Brian: How do you design and build your lab automation & robotics systems?
Austin: We have a significant automation team. they work with the lab scientists to determine the capabilities needed to run the workflows and assemble the right automation systems. we have different types of automation depending on the use case. usually we're using mostly off the shelf components so are primarily focused on the integration of a number of hardware systems along with the software to run it. most of our automation has designed to be relatively flexible and modular.
Brian: Do you have an internal digital platform to manage workflows in your lab automation systems?
Brian: Currently lab instruments are mostly designed for humans.
do you see industry standards developing for lab automation and lab instruments? ie instrument api's for robotics, data, workflow
Austin: Yes we have a number of software to manage workflows and part of the problem is that there isn't a standard and therefore we have more than one. there's been little push to make instrument vendors open up apis let alone standardize them.
Luis: https://sila-standard.com/|https://sila-standard.com/
Question 11:
Apoorva: Thanks for the q&a! what were some of the biggest integration challenges post zymergen acquisition?
Austin: Zymergen was tackling similar problems as ginkgo but many times the approach to the solution would be quite different. figuring out how to integrate the different systems can be a challenge.
Question 12:
Joseph: Do you think using alphafold will become a main part of your pipeline for protein engineering problems?
Austin: Alphafold will be one piece of the toolkit for protein engineering but will unlikely take over everything. we go after a number of different types of problems and want to make sure we have access to a wide range of tools.
Question 13:
Karthik: What are the areas you think should public cloud vendors like aws can improve to accelerate research and development efforts?
Austin: Aws already does quite a bit in the biology space and there's nothing immediate that comes to mind for new areas.
Karthik: Thankyou!
Question 14:
Nicholas: What are some of the trends in the software/ml world that you are most excited about, especially with respect to how they can improve synthetic biology?
Austin: One area is about why we need software programmers. software is about as predictable a system as you can get buty maybe soon we will have ai writing all of our code. once ai can understand programming computers, maybe they can be tasked on to programming dna.
Question 15:
Nicholas: What advice would you give to people thinking about starting their own company?
Austin: Do it. what are you waiting for? the worst that can happen is you fail and go do something else. or you could change the world and become a billionaire.
Tim: +1 for “do it”. starting a company is not easy, yet it’s an exciting journey full of various obstacles and inexplicable miracles.
and be ready that things can be overwhelming sometimes, you will need to put things in order from time to time to resolve that overwhelming feeling.
Nicholas: We’re running up at the time limit, but a big thanks to @austin for hopping on this channel and answering all of our questions!